Subject: Re: Who holds the copyright?
From: "Stephen J. Turnbull" <>
Date: Thu, 18 Nov 1999 16:37:06 +0900 (JST)

>>>>> "Brian" == Brian Behlendorf <> writes:

    Brian> On Wed, 17 Nov 1999, Russell Nelson wrote:
    >> [1] Anything, that is, with a small enough set of copyright
    >> holders to actually *agree* to change a license.

    Brian> Even then, it's not clear.  E.g., when people submit a
    Brian> patch to Linus, are they granting Linus the same rights to
    Brian> that code as if it were his own?  E.g., to relicense?  Keep

Interesting question.  What does happen with jointly authored works in 

For patches small enough to be considered "fair use", apparently.  I
guess the FSF has a rule based on a small number of lines as to
whether they demand an assignment before applying it.  Evidently their 
lwayers consider that they can use, license, and defend the copyright
on such patches without assignment.

But that wouldn't apply to other people's drivers and stuff.

    Brian> in mind Linus has declared that he doesn't consider a
    Brian> closed-source driver to be a violation of the GPL - a right
    Brian> he can only grant if in essence everyone who has
    Brian> contributed patches to him also agrees with this,
    Brian> especially since it's not even written down in the kernel's
    Brian> own copyright notice.

There's an alternative interpretation.  And that is that the fact that
you can apply ld to a couple of compiled objects to create a single
executable does not mean that the source files for the objects are
part of the same work.  AFAIK the definition of work (sources that
generate objects that get exec'ed as a unit are part of the same work)
has not been tested in court.

I've asked before (but still not gotten an answer I understand) where
the `exec' function got the power to define a `work'.

And of course this doesn't apply to kernel modules (either separately
linked and insmod'ed or integrally linked), the exec interface is
irrelevant to them.

    Brian> I would hope that if it ever went to court, the courts
    Brian> would find that when someone submits a patch to the
    Brian> copyright owners, absent any other legal contract, there'd
    Brian> be an implied contract granting copyright.  But who knows.

RMS presumably has gotten sufficient advice to have a pretty good idea,
and he's obviously not comfortable without assignments.

Note that some of the "objectionable" language in the not-quite-good-
enough wannabe-free licenses from companies like Sun and Netscape and
IBM is demanding certain rights over derived work.  One may deduce
that their lawyers are worried about this problem.

University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
What are those two straight lines for?  "Free software rules."